US7944743B2

US7944743B2 - Methods of making a semiconductor memory device

Info

Publication number: US7944743B2
Application number: US12/537,470
Authority: US
Inventors: Sanh D. Tang; Gordon A. Haller; Daniel H. Doyle
Original assignee: Micron Technology Inc
Current assignee: US Bank NA
Priority date: 2006-09-07
Filing date: 2009-08-07
Publication date: 2011-05-17
Anticipated expiration: 2026-09-07
Also published as: TWI353041B; US8446762B2; TW200822292A; US7589995B2; KR20090040460A; US20080061346A1; WO2008030588A1; KR101057204B1; US20110171802A1; US20090311845A1

Abstract

One-transistor (1T) capacitor-less DRAM cells each include a MOS transistor having a bias gate layer that separates a floating body region from a base substrate. The MOS transistor functions as a storage device, eliminating the need of the storage capacitor. Logic “1” is written to and stored in the storage device by causing majority carriers (holes in an NMOS transistor) to accumulate and be held in the floating body region next to the bias gate layer, and is erased by removing the majority carriers from where they are held.

Description

RELATED PATENT DATA

This patent resulted from a divisional application of U.S. patent application Ser. No. 11/516,814, filed Sep. 7, 2006, now U.S. Pat. No. 7,589,995, entitled “One-Transistor Memory Cell with Bias Gate”, naming Sanh D. Tang, Gordon A. Haller and Daniel H. Doyle as inventors, the disclosure of which is incorporated by reference.

TECHNICAL FIELD

This document relates generally to semiconductor integrated circuit technology and particularly, but not by way of limitation, to one-transistor (1T) capacitor-less memory device in which the transistor includes a bias gate separating a floating body from the base substrate of the device.

BACKGROUND

A known dynamic random access memory (DRAM) cell includes a transistor and a capacitor. The transistor functions as a switch controlling whether a data bit is being written into, read from, or held in the DRAM cell. The capacitor functions as the storage device. This one-transistor/one-capacitor (1T/1C) structure limits the extent to which the DRAM cell can be miniaturized and hence the memory capacity of the DRAM device given a certain physical size. The increasing need for smaller electronic systems and larger memory capacity (such as multi-gigabytes), among other reasons, requires reduction in size of the physical structures inside a memory device. While the minimum size of the transistor has been shrinking with the advance of the semiconductor fabrication technology, the size of the capacitor has become the bottleneck in miniaturization of the DRAM cell.

A capacitor-less DRAM cell technology has been provided by fabricating a metal-oxide semiconductor field-effect transistor (MOSFET) on a silicon-on-insulator (SOI) wafer. Because of the insulator between the MOSFET and the base substrate of the wafer, the body of the MOSFET is electrically floating. This effect is utilized to store data by storing a charge (holes in an n-channel MOSFET, or NMOS) in and drawing the charge out of the floating body, which performs the functions of the capacitor in the typical DRAM cell. Such a “floating body cell”, or FBC, eliminates the need for the capacitor in a DRAM cell, thereby removing the capacitor as a bottleneck in the miniaturization of the DRAM cell. However, SOI devices are more costly to produce and therefore have been used primarily for high-end applications. Thus, to increase memory capacity without substantially increasing the size and the cost of DRAM devices, there is a need for reducing the size of a DRAM cell in a less expensive way.

BRIEF DESCRIPTION OF THE DRAWINGS

The drawings illustrate generally, by way of example, but not by way of limitation, various embodiments discussed in the present document.

FIGS. 1A-D are illustrations of an embodiment of a structure of portions of a DRAM device showing a MOS transistor having a floating body formed on a bias gate during various states of operation.

FIG. 2 is an illustration of a specific embodiment of the structure of portions of the DRAM device of FIG. 1.

FIG. 3 is an illustration of another specific embodiment of the structure of portions of the DRAM device of FIG. 1.

FIG. 4 is a circuit schematic illustrating DRAM cells using transistors of FIG. 1.

FIG. 5 is a block diagram illustrating an embodiment of a memory device utilizing the transistors of FIG. 1.

FIG. 6 is a block diagram illustrating an embodiment of a processor-based system utilizing the transistors of FIG. 1.

FIGS. 7A-C are illustrations of shallow-trench-isolation (STI) and active area formation in an embodiment of a fabrication process for a DRAM device.

FIGS. 8A-C are illustrations of silicon (Si) island formation in the embodiment of the fabrication process.

FIGS. 9A-C are illustrations of bias gate formation in the embodiment of the fabrication process.

FIGS. 10A-C are illustrations of planarization in the embodiment of the fabrication process.

FIGS. 11A-C are illustrations of gate dielectric formation in the embodiment of the fabrication process.

FIGS. 12A-C are illustrations of array recessed gate formation in the embodiment of the fabrication process.

FIGS. 13A-C are illustrations of source/drain poly plug contacts formation in the embodiment of the fabrication process.

FIGS. 14A-C are illustrations of common source interconnection formation in the embodiment of the fabrication process.

FIGS. 15A-C are illustrations of array drain and peripheral local interconnection formation in the embodiment of the fabrication process.

DETAILED DESCRIPTION

In the following detailed description, reference is made to the accompanying drawings which form a part hereof, and in which is shown by way of illustration specific embodiments in which the invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention, and it is to be understood that the embodiments may be combined, or that other embodiments may be utilized and that structural, logical and electrical changes may be made without departing from the spirit and scope of the present invention. References to “an”, “one”, or “various” embodiments in this disclosure are not necessarily to the same embodiment, and such references contemplate more than one embodiment. The following detailed description provides examples, and the scope of the present invention is defined by the appended claims and their legal equivalents.

In this document, a “MOS transistor” or “MOSFET” refers to a metal-oxide semiconductor field-effect transistor, an “NMOS transistor” refers to an n-channel metal-oxide semiconductor field-effect transistor (or n-channel MOSFET), and a “PMOS” refers to a p-channel metal-oxide semiconductor field-effect transistor (or p-channel MOSFET).

In this document, “logic ‘1’” and “logic ‘0’” refer to binary digits represented by two different voltage levels in a digital circuit. “Logic ‘1’” is represented by the higher voltage of the two different voltage levels. “Logic ‘0’” is represented by the lower voltage of the two different voltage levels.

This document discusses one-transistor (1T) capacitor-less DRAM cells. “Capacitor-less” (also referred to as zero-capacitor or 0C) refers to the lack of a structure intentionally formed as a storage capacitor in a memory cell. Each 1T capacitor-less DRAM cell includes a MOS transistor having a bias gate layer that separates a floating body region from a base substrate. The MOS transistor functions as a storage device, eliminating the need of the storage capacitor. Logic “1” is written to and stored in the storage device by causing majority carriers (holes in an NMOS transistor) to accumulate and be held in the floating body next to the bias gate layer. Logic “1” is erased (i.e., logic “0” is written) by removing the majority carriers from where they are held.

FIGS. 1A-D are illustrations of an embodiment of a structure of portions of a DRAM device 100 showing a MOS transistor 104 during various states of operation. DRAM device 100 includes 1T capacitor-less memory cells each formed with one MOS transistor 104. MOS transistor 104 is formed on a base substrate 102 made of single crystalline silicon. A conductive bias gate layer 106 is formed on a gate dielectric 105 over base substrate 102. The bias gate layer 106 is made of in situ doped (ISD) polysilicon. A floating body region 108 is formed on a gate dielectric 107 over bias gate layer 106 and coupled between a source region 110 and a drain region 112. Source region 110 separates a source terminal 116 from floating body region 108. Drain region 112 separates a drain terminal 118 from floating body region 108. A gate 114 is formed on floating body region 108 between source region 110 and drain region 112 and is separated from source region 110, drain region 112, and floating body region 108 by a gate dielectric 115. Gate 114 is coupled between a gate terminal 120 from gate dielectric 115. An oxide 122 forms an isolation surrounding MOS transistor 104.

In one embodiment, bias gate layer 106 is formed after the formation of floating body region 108. Bias gate layer 106 is self-aligned to floating body region 108 as the conductive polysilicon is deposited onto base substrate 102 between the formed floating body regions 108. In one embodiment, base substrate 102 and floating body region 108 are formed by a single crystalline silicon, and

gate dielectrics

105, 107, and 115 include the same oxide formed at the same time during the fabrication process. An example of a process for fabricating MOS transistor 104 is discussed below, with reference to FIGS. 7-15. In FIGS. 2-3 and 9-15, a gate dielectric exists between the gate and the source, drain, and floating body, between the bias gate and the floating body, and between the bias gate and the base substrate (though not explicitly shown or labeled).

During operation, source terminal 116 is coupled to a ground potential (0 V). Bias gate layer 106 is coupled to a negative potential. In one embodiment, bias gate layer 106 is coupled to a negative potential between approximately −0.25 V to −2.0 V. In a specific embodiment, bias gate layer 106 is coupled to a negative potential of approximately −1.0 V.

FIG. 1A illustrates an embodiment of the operation for writing logic “1”. To write logic “1”, gate terminal 120 is coupled to a negative gate potential for writing, and drain terminal 118 is coupled to a positive drain potential for writing. In one embodiment, the negative gate potential and positive drain potential create a gate-induced drain leakage (GIDL) current. Majority carriers (holes) 124 are attracted to the negative potential of bias gate layer 106, such that logic “1” is written into MOS transistor 104. In one embodiment, the negative gate potential for writing is between approximately −1.5 V to −3.0 V, and the positive drain potential for writing is between approximately 0.8 V to 2.0 V. In a specific embodiment, the negative gate potential for writing is approximately −2.5 V, and the positive drain potential for writing is approximately 1.8 V. The GIDL current is created by the high gate-to-drain voltage (V_GD=4.3 V).

In another embodiment, to write logic “1”, gate terminal 120 is coupled to a positive gate potential for writing, and drain terminal 118 is coupled to a positive drain potential for writing. A lower positive gate potential and a higher positive drain potential creates impact ionization. MOS transistor 104 operates in the current saturation region, causing a current to flow while allowing impact ionization to occur near drain region 112. The impact ionization produces holes, which are attracted to the negative potential of bias gate layer 106. In one embodiment, the positive gate potential for writing is between approximately 0.7 V to 1.5 V, and the positive drain potential for writing is between approximately 1.0 V to 2.5 V. In a specific embodiment, the positive gate potential for writing is approximately 1.0 V, and the positive drain potential for writing is approximately 1.5 V.

FIG. 1B illustrates an embodiment of the operation for writing logic “0”, i.e., erasing logic “1”. In the embodiment of impact ionization, to write logic “0”, gate terminal 120 is coupled to a positive gate potential for erasing, and drain terminal 118 is coupled to a negative drain potential for erasing. The inverted channel and the negative drain potential create a potential that removes holes 124 from the region of floating body region 108 near bias gate layer 106. In one embodiment, the positive gate potential for erasing is between approximately 0.7 V to 1.5 V, and the negative drain potential for erasing is between approximately −0.3 V to −1.0 V. In a specific embodiment, the positive gate potential for erasing is approximately 1.0 V, and the negative drain potential for erasing is approximately −1.0 V. In the embodiment of GIDL current, to write logic “0”, gate terminal 120 is coupled to a negative gate potential for erasing, and drain terminal 118 is coupled to a positive drain potential for erasing. In one embodiment, the negative gate potential for erasing is between approximately −1.5 V to −3.0 V, and the positive drain potential for erasing is between approximately 0.0 V to 0.5 V. In a specific embodiment, the negative gate potential for erasing is approximately −1.8 V, and the positive drain potential for erasing is approximately 0.0 V (ground potential). In one embodiment, the use of GIDL current is limited to refresh or read-modify-write operations because the terminal potentials given above may not overwrite logic “1” with a logic “0”.

In one embodiment, the gate potentials for writing logic “1” and logic “0” (i.e., for writing and erasing) are substantially the same. Because random data consisting of logic “1's” and “0's” are expected to be written, using the same gate potential avoids limiting the speed of operation due to the time required for changing the gate potential.

FIG. 1C illustrates an embodiment of the operation of reading. To read, gate terminal 120 is coupled to a positive gate potential for reading, and drain terminal 118 is connected to a positive drain potential for reading. When holes 124 are present in floating body region 108 near bias gate layer 106, a high gate current results in a logic “1” reading. When holes 124 are not present in floating body region 108 near bias gate layer 106, a low gate current results in a logic “0” reading. In one embodiment, the positive gate potential for reading is between approximately 0.5 V to 1.5 V, and the positive drain potential for reading is between approximately 0.2 V to 1.0 V. In a specific embodiment, the positive gate potential for reading is approximately 1.0 V, and the positive drain potential for reading is approximately 0.3 V.

FIG. 1D illustrates an embodiment of the operation of reading. To hold, gate terminal 120 is coupled to a negative gate potential for holding, and drain terminal 118 is coupled to a positive drain potential. When logic “1” has been written, holes 124 are confined in floating body region 108 near bias gate layer 106. In one embodiment, the negative gate potential for holding is between approximately −0.2 V to −1.0 V, and the positive drain potential for reading is between approximately 0.0 V to 0.5 V. In a specific embodiment, the negative gate potential for reading is approximately −0.5 V, and the positive drain potential for reading is approximately 0 V (the ground potential).

FIG. 2 is an illustration of an embodiment of the structure of portions of a DRAM device 200, which represents a specific embodiment of DRAM device 100. DRAM device 200 includes 1T capacitor-less memory cells each formed with a MOS transistor 204. MOS transistor 204 is a specific embodiment of MOS transistor 104 and is implemented as a buried recessed-access device (RAD).

MOS transistor

204 is formed on base substrate 102 between oxide isolations 122. MOS transistor 204 includes bias gate layer 106 formed on base substrate 102, floating body region 108 on bias gate layer 106 between source region 110 and drain region 112, and buried gate dielectric 214 between source region 110 and drain region 112. Source terminal 116 is connected to a ground potential line 230. Gate terminal 120 is connected to a word line 232. Drain terminal 118 is connected to a bit line 234 via contact plug 236.

FIG. 3 is an illustration of an embodiment of the structure of portions of a DRAM device 300, which represents a specific embodiment of DRAM device 100. DRAM device 300 includes 1T capacitor-less memory cells each formed with a MOS transistor 304. MOS transistor 304 is a specific embodiment of MOS transistor 104 and is implemented as a strapped RAD.

MOS transistor

304 is formed on base substrate 102 between oxide isolations 122. MOS transistor 304 includes bias gate layer 106 formed on base substrate 102, floating body region 108 on bias gate layer 106 between source region 110 and drain region 112, and buried gate dielectric 314 between source region 110 and drain region 112. Source terminal 116 is connected to a ground potential line 330 via contact plug 336. Gate terminal 118 is connected to a word line 332. Drain terminal is connected to a bit line 334.

FIG. 4 is a circuit schematic illustrating a portion of a DRAM device 400 including DRAM cells 410. DRAM cells 410 are 1T capacitor-less memory cells each include a MOS transistor 404. In one embodiment, MOS transistor 404 is MOS transistor 104. In specific embodiments, MOS transistor 404 is MOS transistor 204 or MOS transistor 304. MOS transistor 404 has a drain terminal coupled to a bit line 406, a gate terminal coupled to a word line 408, and a source terminal coupled to a ground potential.

FIG. 5 is a block diagram illustrating an embodiment of a memory device 520 utilizing MOS transistor 104 (including its specific embodiments such as MOS transistors 204 or 304). Memory device 520 contains a memory array 522, a row decoder 524, a column decoder 528, and sense amplifiers 526. Memory array 522 includes a plurality of transistor cells 540 each being a 1T capacitor-less memory cell formed with a MOS transistor such as MOS transistor 104. Word lines 538 and bit lines 530 are commonly arranged into rows and columns. Bit lines 530 of memory array 522 are connected to sense amplifiers 526, while word lines 538 are connected to row decoder 524. Address and control signals are input on address/control lines 532 into memory device 520 and are connected to column decoder 528, sense amplifiers 526, and row decoder 524, and are used to gain read and write access, among other things, to memory array 522.

Column decoder

528 is connected to sense amplifiers 526 via control and column select signals on column select lines 534. Sense amplifiers 526 receives input data destined for memory array 522, and outputs data read from memory array 522 over input/output (I/O) data lines 536. Data is read from the cells of memory array 522 by activating one of the word lines 538 (via the row decoder 524), which couples all of the memory cells corresponding to that word line to respective bit lines 530, which define the columns of the array. One or more bit lines 530 are also activated. When a particular word line 538 and bit lines 530 are activated, sense amplifiers 526 connected to a bit line column detects and amplifies the conduction sensed through a given transistor cell and transferred to its bit line 530 by measuring the potential difference between the activated bit line 530 and a reference line which may be an inactive bit line. In the read operation the source region of a given cell is coupled to bit line. The operation of Memory device sense amplifiers is described, for example, in U.S. Pat. Nos. 5,627,785; 5,280,205; and 5,042,011, all assigned to Micron Technology Inc., the specification of which are incorporated by reference herein in their entirety.

FIG. 6 is a block diagram illustrating an embodiment of a processor-based system utilizing MOS transistor 104 (including its specific embodiments such as MOS transistors 204 or 304). FIG. 6 is a block diagram of an electrical system, or processor-based system, 600 utilizing 1T capacitor-less memory cells each formed with a MOS transistor such as MOS transistor 104. The processor-based system 600 may be a computer system, a process control system or any other system employing a processor and associated memory. System 600 includes a central processing unit (CPU) 652, e.g., a microprocessor that communicates with the memory 662 and an I/O device 658 over a bus 690. It is noted that bus 690 may be a series of buses and bridges commonly used in a processor-based system, but for convenience purposes only, bus 690 has been illustrated as a single bus. A second I/O device 660 is illustrated, but is not necessary. The processor-based system 600 can also include read-only memory (ROM) 664 and may include peripheral devices such as a floppy disk drive 654 and a compact disk (CD) ROM drive 656 that also communicates with the CPU 652 over the bus 690 as is well known in the art.

It will be appreciated by those skilled in the art that additional circuitry and control signals can be provided, and that the processor-based system 600 has been simplified to help focus on the invention.

FIG. 6 illustrates an embodiment for electronic system circuitry in which one or more transistors including those as illustrated in FIGS. 1A-D, 2, and 3 are used. The illustration of system 600, as shown in FIG. 6, is intended to provide a general understanding of one application for the structure and circuitry of the present invention, and is not intended to serve as a complete description of all the elements and features of an electronic system using the novel 1T capacitor-less memory cells discussed in this document. Further, the invention is equally applicable to any size and type of system 600 using the one or more transistors including those as illustrated in FIGS. 1A-D, 2, and 3, and is not intended to be limited to that described above. As one of ordinary skill in the art will understand, such an electronic system can be fabricated in single-package processing units, or even on a single semiconductor chip, in order to reduce the communication time between the processor and the memory device.

Applications containing MOS transistors each having a floating body region separated from a base substrate by a conductive biased gate layer, such as MOS transistor 104 and its specific embodiments described in this document, include electronic systems for use in memory modules, device drivers, power modules, communication modems, processor modules, and application-specific modules, and may include multilayer, multichip modules. Such circuitry can further be a subcomponent of a variety of electronic systems, such as a clock, a television, a cell phone, a personal computer, an automobile, an industrial control system, an aircraft, and others.

FIGS. 7-15 illustrate an embodiment of a fabrication process for a DRAM device according to the present subject matter. FIGS. 7-15 and the associated description below illustrate the process of making MOS transistors having floating body regions separated from a base substrate by a bias gate layer, such as

MOS transistors

104, 204, and 304 as discussed above. While not intended to be a full description of the fabrication process, FIGS. 7-15 and the associated description below provide sufficient details to enable those skilled in the art to make 1T capacitor-less memory devices by practicing the invention.

The size of each physical structure of a memory device is described by the size of electrically conductive lines (word and bit lines) in terms of lithographic feature size (F). The lithographic feature size (F) is one half of the minimum pitch, i.e., one half of the sum of the width of one of the electrically conductive lines and the width of the isolation space between the electrically conductive lines. An 8F²memory cell refers to a memory cell that has an area of 8 square lithographic features. Specific sizes discussed below with references to FIGS. 7-15 are based on a process for fabricating of 8F²memory cells. However, the process illustrated in FIGS. 7-15 is generally applicable to fabricate memory cells with other cell sizes.

FIGS. 7A-C illustrate the formation of shallow-trench-isolations (STIs) 702 and active areas between the STIs 702. FIG. 7A is a top view of the DRAM device being fabricated, and FIGS. 7B-C are each a cross-sectional view. Shallow-trench-isolations 702 and the active areas are formed on a silicon base substrate 700. The active areas are each 3F wide. STIs 702 are each 1F wide and made of silicon oxide (SiO₂). The depth of each STI 702 is about 3,500 angstroms (Å). A nitride cap 704 is formed on each active area.

FIGS. 8A-C illustrate the formation of silicon islands 810. FIG. 8A is a top view of the DRAM device being fabricated, and FIGS. 8B-C are each a cross-sectional view. Silicon islands 810 are formed using a pattern of 1F lines separated by 1F spaces and perpendicular to the STIs 702 and the active areas. An in situ trench etch is performed to remove both STIs 702 and the active areas along the 1F lines with a depth of about 2,000 angstroms. An in situ undercut etch is then performed to form the silicon islands 810 by disconnecting them from base substrate 700. Silicon islands 810 are held by STIs 702. Nitride cap 704 also helps prevent silicon islands 810 from being broken off.

FIGS. 9A-C illustrate the formation of a conductive bias gate layer 920. FIG. 9A is a top view of the DRAM device being fabricated, and FIGS. 9B-C are each a cross-sectional view. With nitride cap 704 intact, the exposed silicon is oxidized to form gate dielectric. In situ-doped (ISD) polysilicon is deposited to fill the gaps between silicon islands 810. A combination of chemical-mechanical polishing (CMP) and reactive ion etching (RIE), or a simple RIE, is performed to recess the ISD polysilicon at about 1,500 angstroms from a surface 924, which is the top surface of the device during the fabrication process. This leaves about 500 angstroms of the ISD polysilicon on the side of silicon islands 810. At this point, conductive bias gate layer 920 is defined and self-aligned to silicon islands 810. Voids 922 may be present in the ISD polysilicon, which do not affect the function of bias gate layer 920.

Bias gate layer

920 separates silicon islands 810 from base substrate 700. The gate dielectric formed on the exposed silicon before the deposition of the ISD polysilicon insulates silicon islands 810, thus making silicon islands 810 the floating bodies of MOS transistors. That is, when the fabrication process for the DRAM device is completed, each silicon island becomes floating body region 108, and each bias gate 920 becomes bias gate layer 106, of a MOS transistor such as

MOS transistor

104, 204, or 304 as discussed above.

FIGS. 10A-C illustrate planarization of surface 924. FIG. 10A is a top view of the DRAM device being fabricated, and FIGS. 10B-C are each a cross-sectional view. The exposed ISD polysilicon of bias gate layer 920 is oxidized using thin oxidation. The trenches are filled with SiO₂, such as by nitride liner and spin-on-dielectric (SID). A stop-on-nitride (SON) CMP is then performed. Surface 924 is thus planarized, with sacrificial nitride exposed for subsequent wet nitride removal.

FIGS. 11A-C illustrate removal of nitride cap 704. FIG. 11A is a top view of the DRAM device being fabricated, and FIGS. 11B-C are each a cross-sectional view. The remaining portions of nitride cap 704 are removed by a hot phosphoric wet nitride removal. Thick or thin gate dielectric is formed by oxidation.

FIGS. 12A-C illustrate the formation of array recessed gates 1230. FIG. 12A is a top view of the DRAM device being fabricated, and FIGS. 12B-C are each a cross-sectional view. In one embodiment, recessed-access-device (RAD) is formed through a polysilicon process. Silicon islands 810 are recessed by trench etch to about 1,000 angstroms below surface 924. The resultant exposed silicon is oxidized to form the gate dielectric. Array and “dummy” gates may be defined at the same time. Dummy (ground) gates 1232 are defined in the array over STIs 702 to isolate the subsequent polysilicon plugs. Gate nitride spacers 1234 are formed on array recessed gates 1230 and dummy gates 1232.

FIGS. 13A-C illustrate the formation of drain and source polysilicon plugs 1340. FIG. 13A is a top view of the DRAM device being fabricated, and FIGS. 13B-C are each a cross-sectional view. Drain and source implant may be done before or after the formation of gate nitride spacers 1234 to enhance the GIDL effects. Drain and source polysilicon plugs 1340 are formed by teos or phosphor-silicon-glass or boro-phosphor-silicon-glass deposition and reflow. A stop-on-nitride (SON) CMP follows. In one embodiment, adjacent source contacts are arranged to tie together to lower contact resistance.

FIGS. 14A-C illustrate the formation of common source interconnections 1450. FIG. 14A is a top view of the DRAM device being fabricated, and FIGS. 14B-C are each a cross-sectional view. A thin oxide is deposited on surface 924, followed by a common source contact etch to open the source contact. A common source interconnect conductor 1452 is deposited. A nitride cap 1454 is formed on the common source interconnect conductor 1452.

FIGS. 15A-C illustrate the formation of array drain and peripheral local interconnections. FIG. 15A is a top view of the DRAM device being fabricated, and FIGS. 15B-C are each a cross-sectional view. Drain interconnects (bit lines) 1560 are each formed with an interconnect conductor 1562, and nitride cap 1564 is deposited on the interconnect conductor 1562. Examples of the interconnect conductor include TiSix/W and metal-mode-titanium/Tungsten (MMTi/W). An example of the material forming the nitride cap is Si₃N₄. Array drain contacts are self-aligned-contact (SAC) etched to expose the polysilicon below. Peripheral contacts of the interconnects are exposed by etching.

The peripheral contacts illustrated in FIG. 15A include source contacts 1570, gate contacts 1572, drain contacts 1574, and bias gate contacts 1756. These contacts are each connected to a potential line that is to be coupled to a specified potential during the operation of the DRAM device. The specified potential may be constant during the operation, such as the potential coupled to the source terminal and the bias gate layer of a MOS transistor of the DRAM device, or variable during the operation, such as the potential coupled to the gate terminal and the drain terminal of the MOS transistor. An example of the potentials to which source contacts 1570, gate contacts 1572, drain contacts 1574, and bias gate contacts 1756 are coupled to during the operation of the DRAM device is given above in FIGS. 1A-D and the associated discussion. In this example, assuming the contacts provide for connections between MOS transistor 104 and the potential lines, source contact 1570 is coupled to source terminal 116, gate contact 1572 is coupled to gate terminal 120, drain contact 1574 is coupled to drain terminal 118, and bias gate contact 1756 is coupled to bias gate layer 106.

This disclosure includes several processes and structures. The present invention is not limited to a particular process order or structural arrangement. Although specific embodiments have been illustrated and described herein, it will be appreciated by those of ordinary skill in the art that any arrangement which is calculated to achieve the same purpose may be substituted for the specific embodiments shown. This application is intended to cover adaptations or variations. It is to be understood that the above description is intended to be illustrative, and not restrictive. Combinations of the above embodiments, and other embodiments, will be apparent to those of skill in the art upon reviewing the above description. The scope of the present invention should be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled.

Claims

1. A method for making a semiconductor memory device, the method comprising:

providing a semiconductor substrate;

forming shallow-trench-isolations in the semiconductor substrate;

forming floating bodies and a base substrate out of the semiconductor substrate by removing portions of the semiconductor substrate and the shallow-trench-isolations, the floating bodies each being a disconnected portion of the semiconductor substrate held separate from the base substrate by portions of the shallow-trench-isolations which were not removed;

forming a conductive layer between the floating bodies and the base substrate; and

forming MOS transistors each including one of the floating bodies,

wherein the floating body of each of the MOS transistors is adapted to store a data bit during operation of the memory device.

2. The method of claim 1, further comprising forming a nitride cap on a surface of the semiconductor substrate between the shallow-trench-isolations before forming the floating bodies, the nitride cap further holding the floating bodies separate from the base substrate.

3. The method of claim 1, wherein the semiconductor substrate comprises a single crystalline silicon.

4. The method of claim 3, wherein the conductive layer is a polysilicon layer.

5. The method of claim 4, wherein forming the conductive layer comprises depositing polysilicon into gaps formed between the floating bodies such that the conductive layer is self-aligned to the floating bodies.

6. A method for making a semiconductor memory device, the method comprising:

providing a semiconductor substrate;

forming isolation regions in the semiconductor substrate;

forming floating bodies and a base substrate out of the semiconductor substrate by removing portions of the semiconductor substrate and the isolation regions, the floating bodies each being a disconnected portion of the semiconductor substrate held separate from the base substrate by portions of the isolation regions which were not removed;

forming conductive material between the floating bodies and the base substrate; and

forming field effect transistors each including one of the floating bodies.

7. The method of claim 6 wherein forming the conductive material comprises forming voids in the conductive material.

8. The method of claim 7 comprising forming the voids underneath the floating bodies.

9. The method of claim 8 wherein the respective voids are received underneath only a portion of the floating bodies.

10. The method of claim 7 comprising forming the voids to respectively extend from one isolation region to another isolation region.

11. The method of claim 7 comprising forming the voids underneath the floating bodies, and to respectively extend from one isolation region to another isolation region.

12. The method of claim 6 comprising forming gates of the field effect transistors to be recessed in the semiconductor substrate.

13. The method of claim 6 comprising forming a dummy gate over the respective isolation regions.

14. The method of claim 13 comprising forming gates of the field effect transistors to be recessed in the semiconductor substrate.

15. The method of claim 6 wherein forming the conductive layer comprises depositing and subsequently recessing conductive material.